ANALYTICAL BIOCHEMISTRY ARTICLE NO.

255, 9094 (1998)

AB972435

Purication of Thymine Glycol DNA and Nucleosides by Use of Boronate Chromatography1

Bozidar Jerkovic, Hsiang Chuan Kung, and Philip H. Bolton
Department of Chemistry, Wesleyan University, Middletown, Connecticut 06459

Received July 10, 1997

Boronate columns can be used to purify DNAs containing cis-thymine glycol residues and can also be used to purify cis-thymine glycol nucleosides. The boronate group can form a reversible complex with the cis-diol of the thymine glycol but not with the urea residue which is formed by alkaline hydrolysis of the thymine glycol. This method is rapid and appears applicable to a range of nucleic acids. In addition to the purications of DNAs and nucleosides demonstrated here boronate chromatography may be applicable to assaying the extent and sites of oxidative damage to DNAs. 1998 Academic Press

The exposure of DNA in cells, bound to proteins, as a solid or free in solution to ionizing radiation or to oxidative stress can lead to the conversion of thymine to thymine glycol (5,6-dihydroxy-5,6-dihydrothymine) (19). Approximately 1020% of the damage to DNA induced by ionizing radiation, including that used to treat tumors, is the result of thymine base oxidation and fragmentation (1, 2, 5, 1013) and specic repair enzymes for thymine glycol have been found in a wide range of organisms. We have recently developed methods for the preparation of single-stranded and duplex DNAs containing a single thymine glycol residue (14) and this work showed that the permanganate oxidation of DNA leads to a single, (5R,6S) cis isomer of thymine glycol. These pure thymine glycol-containing DNAs have been used in structural (1416) and biochemical studies (Scheme 1) (17, 18). While we have been able to develop a method for preparing pure samples of DNA containing a single thymine site, further development was needed. The methods we have been using for the oxidation of the
1 This work was supported by American Cancer Society Grant NP750.

DNA lead to a wide range of products. The identication of the thymine glycol containing portions of the reaction mixture has relied on the digestion of the DNA to nucleosides followed by chromatography to nd the thymine glycol nucleoside peak. A physical method for separating the thymine glycol containing portion of the reaction mixture could be very useful when new DNA sequences are studied. There is growing interest in being able to quantify the extent of oxidative damage of the DNA extracted from tumor and other cells. The amount of thymine glycol present is one marker of the extent of oxidative damage. To be able to ascertain the amount of thymine glycol present, as nucleoside, nucleotide, or in DNA, a purication method that relies on the unique physical properties of thymine glycol is needed to be able to reliably separate thymine glycol-containing nucleic acid from a mixture that may contain a variety of oxidized DNAs. The use of boronate columns to purify thymine glycol-containing DNAs was examined since boronate is known to form a strong, reversible complex with cisdiols (19 22). Also, it was shown sometime ago that the chromatographic mobility of thymine glycol nucleosides was altered upon the addition of boron buffer (13). Thymine glycol is a six-membered ring cis-diol and boronate binds more strongly to ve-membered ring cis-diol (20) but the anticipation was that conditions could be found under which purication could take place. To examine binding specicity we have used both thymine glycol nucleoside and thymine glycol-containing DNA.
MATERIALS AND METHODS

Boronate column chromatography. About 250 mg of Af-gel 601 (Bio-Rad Laboratories, Hercules, CA) was placed in Bio-Spin disposable chromatography columns (Bio-Rad) and hydrated overnight in 100 mM ammo0003-2697/98 $25.00 Copyright 1998 by Academic Press All rights of reproduction in any form reserved.

90

THYMINE GLYCOL PURIFICATION ON BORONATE COLUMNS

91

be used several times before any decrease in binding capacity was observed. Sample preparation. DNA containing thymine glycol was prepared by oxidation of the parent singlestranded DNA with 0.1 M KMnO4 . The oxidation was carried out in a 300-ml plastic jar containing 20 ml of 0.2 M KHPO4 at pH 8.6 and 50 OD (OD 50 mg/ml of double-stranded DNA measured as an absorbance at a wave length of 254 nm) of d(C1G2C3G4A5T6A7C8G9C10C11). The mixture was stirred with a magnetic stirrer for 20 min in an ice bath. The sample at 4 C was then treated with 8 ml of 0.1 M KMnO4 (Aldrich Chemical Co., Inc., Milwaukee, WI) for 5 min. The reaction was quenched by the addition of 0.5 ml of allyl alcohol (Aldrich) which converts the MnO4 into MnO2 . The sample was then kept at 4 C for at least 1 h to allow the MnO2 to completely precipitate. The reaction mixture was then centrifuged to remove the MnO2 and the supernatant containing the products was diluted to 100 ml with distilled water and was desalted by the use of a Waters C18 cartridge (Waters Corp., Milford, MA). The DNA was eluted from the C18 column with 14 ml of 70% CH3CN/H2O in four steps, 3 1 4 ml and 1 1 2 ml. Thymine glycol nucleoside was prepared following the same procedure as for the single-stranded DNA with the exception that 60.5 mg of thymidine nucleoside was used (Sigma Chemical Co., St. Louis, MO) per reaction. After oxidation the nucleoside sample was kept at 4 C overnight instead of for 1 h as was the case for the DNA sample. Urea-containing DNA was prepared by alkaline hydrolysis of thymine glycol-containing single-stranded DNA in 0.025 M sodium phosphate buffer at pH 13.0. The reaction was started by mixing 15 OD of dry DNA with buffer. The reaction was run for 1 h at room temperature after which 11 ml of 0.2 M EDTA at pH 7.0 was added to quench the reaction. The reaction mixture was then desalted as described above. Urea nucleoside was prepared by alkaline hydrolysis of the thymine glycol nucleoside using the same procedure as for DNA. Purification of the oxidized DNA. The single stranded DNA containing thymine glycol was puried by HPLC on a semipreparative reversed-phase PRP-1 column (Hamilton Co., Reno, NV) and eluted, elution time 20.5 min, using a 520% gradient of 50% acetonitrile, 25 mM phosphate buffer at pH 7.0 mixed with 25 mM phosphate buffer at pH 7.0. A typical chromatogram is shown in Fig. 1. The column ow rate was 2 ml/min, and detection was at 254 nm. The other peaks produced by oxidation were not identied and are likely to contain oxidized guanosine, adenosine, and other oxidation products. The puried single-strand DNA containing thymine glycol was collected, diluted, and de-

SCHEME 1

nium bicarbonate, 15 mM MgCl2 , pH 8.7, application buffer, at room temperature. The next day the column was allowed to equilibrate with another 510 ml of application buffer prior to actual use, during which the ow rate was set to 0.40.5 ml/min. All samples were prepared so that a small volume of the sample solution, 20150 ml, was added to 12 ml of application buffer and the resulting mixture was gently shaken and then applied onto the column. This avoided exposure of the samples to concentrated buffers which may have caused partial hydrolysis of the thymine glycol residues. This also allowed for a consistent method of preparing the samples before chromatography. After applying the sample, the volume of 5 ml of application buffer was allowed to pass through, which washed out all other components of the sample from the column except those containing thymidine glycol. The recovery of thymidine glycol from the column is accomplished by applying the elution buffer that contained 100 mM ammonium formate at pH 4.0. Between 10 and 12 ml of elution buffer was allowed to pass through the column and collected. After that the column was regenerated with 15 ml of application buffer for another use. The binding capacity of the columns appeared to be quite sensitive to ow rate. We observed very little binding with a ow rate of 1 ml/min or greater. The binding capacity appeared constant over the range of ow rates between 0.5 and 0.3 ml/min. Reducing the ow rate to less than 0.3 ml/min did not increase the binding capacity. A substantial volume decrease in the boronate gel material was observed after applying elution buffer. The volume change is due to the relatively low pH, 4.0, of the buffer. Although low pH buffers required additional time for the resuspending and repacking of the column, these were chosen over high pH elution buffer as the literature indicated that better resolution and binding are obtained with low pH buffers (20). When we did not intend to use the column material for several days the column was washed in 15 ml of buffer, 100 mM ammonium bicarbonate, 15 mM MgCl2 at pH 7.0 and stored in a refrigerator. Columns could

92

JERKOVIC, KUNG, AND BOLTON

salted using a C18 column. The isolated yield was typically about 30%. Thymine glycol nucleoside was puried using the same procedure. Characterization of the thymine glycol in the singlestranded DNA. To conrm that the DNA prepared as described above contains a single thymine glycol and no other modications, further analysis was carried out. A sample of 1 OD of the single-stranded DNA containing thymine glycol was incubated with 2 units of venom phosphodiesterase, 4.1 units of alkaline phosphatase, 1 M Tris, and 1 M MgCl2 at pH 8.0 and 37 C for 36 h. The resulting nucleoside mixtures were analyzed by HPLC (Fig. 3). This mixture was also used for boronate column chromatography. The HPLC was performed using a PRP-1 column with a 520% gradient of 50% acetonitrile and 25 mM phosphate buffer at pH 7.0 mixed with 25 mM phosphate buffer at pH 7.0. The column ow rate was 2 ml/ min, and the nucleosides were detected at 254 nm. The order of elution of nucleosides is C, G, T, A, with elution times of 13.6, 18.5, 20.00, and 22.9 min, respectively. A typical chromatogram is shown in Fig. 1. The relative ratios of C, G, and A from the thymine glycol-containing DNA were the same as in the parent strand. Thymine glycol elutes at about the same time as T. The analysis of the stereochemistry of thymine glycol produced in this manner has been discussed previously (14, 16, 21).
RESULTS AND DISCUSSION

FIG. 1. The HPLC traces of the mixture of nucleosides containing thymidine glycol, Tg, before (top) and after (middle) boronate column chromatography are shown. The HPLC trace of the mixture after base hydrolysis and boronate column chromatography is shown at the bottom.

Boronate columns can be used to separate thymidine glycol from other nucleosides. The HPLC chromatogram obtained with a mixture of nucleosides is shown in Fig. 1. When the mixture is rst chromatographed over a boronate column and the fraction that binds applied to the HPLC column, a single peak is observed as shown in Fig. 1. This single peak is the cis-thymidine glycol. Base hydrolysis converts the thymine glycol to urea. When the mixture of nucleosides containing thymine glycol is rst base hydrolyzed and then submitted to boronate and HPLC chromatography the HPLC trace is essentially at as shown in Fig. 1. This conrms that only the thymine glycol residues bind to the boronate column. These results show that passage over a boronate column allows the one-step purication of thymine glycol nucleosides. This form of HPLC does not separate (5S,6R) from (5R,6S) thymine glycol, a separation which can be performed by other HPLC methods (15, 16, 22). Thus, boronate columns may be very useful in the purication of thymidine glycol on a large scale for the preparation of phosphoramidites and are also likely to be applicable to mono-, di-, and triphosphate nucleo-

tides. Boronate columns could also be used for the purication of thymidine glycol from mixtures of nucleosides obtained from nuclease digestion of cellular or other DNA to detect the amount of oxidative damage. Since thymine glycol is excreted in urine (22, 23), boronate columns could be part of a noninvasive screen for assaying the amount of thymine oxidative damage. Boronate columns can also be used to aid the purication of DNAs containing thymine glycol residues. Figure 2 shows the HPLC chromatograms of the mixture of oxidized DNAs obtained by permanganate oxidation before and after the mixture had been puried by boronate chromatography. There are a number of peaks observed which correspond to the various forms of oxidized thymine that are produced in the reaction. The results show that the boronate column isolates the thymine glycol-containing DNAs. The hydrolysis of the thymine glycol in the DNA to urea prior to the boronate chromatography abolished binding. The DNAs that were bound to the boronate column were digested to nucleosides and the base composition indicated that the A, G, and C ratios are basically the same as before oxidation with the largest change being the presence of thymine glycol as shown in Fig. 3. This mixture appears to also contain minor amounts of other forms of oxidized nucleosides as indicated by the nonat base-

THYMINE GLYCOL PURIFICATION ON BORONATE COLUMNS

93

line. Base hydrolysis of this mixture followed by boronate and HPLC chromatography gave a at HPLC trace. Thus, the boronate columns allow the separation of cis-thymine glycol-containing DNAs from those without this damaged residue. The boronate columns can be used to aid in the purication of thymine glycolcontaining DNAs for structural and biochemical studies. The boronate columns could also be used, for example, with a mixture of restriction fragments to separate the thymine glycol-containing DNAs which could then be identied by two-dimensional gel electrophoresis or

FIG. 3. The HPLC traces of the mixture of nucleosides formed by the digestion of the thymine glycol-containing DNA before (top) and after (middle) boron column chromatography are shown. The HPLC trace of the mixture after hydrolysis with base and subjected to boronate column chromatography is shown at the bottom.

another method. This could allow the direct and rapid analysis of specic sites of oxidative damage. Boronate columns have been shown to have a high utility in the purication of nucleosides and DNAs that contain thymine glycol. The hydrolysis of thymine glycol to urea provides a readily accessible control sample. The boronate columns have many potential applications in assaying the extent and location of oxidative damage in DNAs.
REFERENCES
1. Basu, A. K., Loechler, E. L., Leadon, S. A., and Essigmann, J. M. (1989) Proc. Natl. Acad. Sci. USA 86, 76777681. 2. Wallace, S. S. (1988) Environ. Mol. Mutagenesis 12, 431477. 3. Ward, J. F. (1988) Prog. Nucleic Acids Res. Mol. Biol. 35, 95 125. 4. Von Sonntag, C. (1987) The Chemical Basis of Radiation Biology, Taylor and Francis, London. 5. Teoule, P. (1987) Int. J. Radiat. Biol. 51, 573589. 6. Freidberg, E. C. (1985) DNA Repair, Freeman, New York. 7. Lett, J. T. (1990) Prog. Nucleic Acids Res. Mol. Biol. 39, 305 352. 8. Iida, S., and Hayatsu, H. (1970) Biochim. Biophys. Acta 213, 1 13. 9. Ekert, B., and Levin, J. D. (1959) Nature 184, BA58BA59.

FIG. 2. The HPLC traces illustrate the steps in the characterization of the products of the single-stranded DNA products that bind to a boronate column. The HPLC trace at the top is of the unmodied single-stranded DNA, and the next uppermost trace is of the oxidation products of the DNA strand. The lower middle trace is of the oxidation products after boronate column chromatography. The bottom trace is of the DNA that had been both base hydrolyzed and subjected to boronate column chromatography before application to the HPLC column.

94

JERKOVIC, KUNG, AND BOLTON 17. Reardon, J. T., Bessho, T., Kung, H. C., Bolton, P. H., and Sancar, A. (1997) Proc. Natl. Acad. Sci. USA 94, 94639468. 18. Hilbert, T. P., Boorstein, R. J., Kung, H. C., Bolton, P. H., Xing, D., Cunningham, R. P., and Teebor, G. W. (1996) Biochemistry 35, 25052511. 19. Lau, H. H., and Baird, W. M. (1994) Carcinogenesis 15, 907915. 20. Schott, H., Rudloff, E., Schmidt, P., Roychoudhury, R., and Kossel, H. (1973) Biochemistry 12, 932938. 21. Weinfeld, M., Soderlind, K.-J. M., and Buchko, G. W. (1993) Nucleic Acids Res. 21, 621626. 22. Ichikawa, N., Watanabe, K., Tomikawa, S., Nagao, T., and Uchida, H. (1996) Transplant. Proc. 28, 17651766. 23. Cao, E. H., and Wang, J. J. (1993) Carcinogenesis 14, 1359 1362.

10. Agarwal, S., and Sohal, R. S. (1994) Proc. Natl. Acad. Sci. USA 91, 1233212335. 11. Ide, H., Kow, Y. W., and Wallace, S. S. (1985) Nucleic Acids Res. 13, 80358052. 12. Hayes, R. C., and LeClerc, J. E. (1986) Nucleic Acids Res. 14, 10451061. 13. Teebor, G. W., Boorstein, R. J., and Cadet, J. (1988) Int. J. Radiat. Biol. 54, 131150. 14. Kao, J. Y., Goljer, I., Phan, T. A., and Bolton, P. H. (1993) J. Biol. Chem. 268, 1778717793. 15. Kung, H. C., Parker, S. A., Wang, K. Y., and Bolton, P. H. (1996) Magn. Reson. Chem. 34, S47S54. 16. Kung, H. C., and Bolton, P. H. (1997) J. Biol. Chem. 272, 9227 9236.